Informal caregivers experience high rates of burden and psychological distress, but whether caregiving has adverse effects on the caregiver's physical health is less well-established. The proposed study aims to a) compare change in physical health between caregivers and non-caregivers over a 2-year period; and b) compare the effects of factors directly associated with caregiving (e.g, hours/day in caregiving activities) and indirectly associated with caregiving (e.g., use of psychotropic medications) on change in physical health over a 2-year period. This study will be conducted as an ancillary study to the Study of Osteoporotic Fractures (SOF), a 4-site prospective study of osteoporosis and related conditions among elderly community-dwelling women. A screening questionnaire will be administered to all eligible SOF participants, including the newly-recruited African-American cohort, at the 6th SOF clinic visit to identify a sample of 600 SOF participants who are informal caregivers (including an estimated 100 African-American caregivers) and 600 non-caregivers, matched on age-, SOF site-, and race. This study will collect data from caregivers and non-caregivers through interviewer-administered questionnaires in face-to-face interviews at baseline and at 1- and 2-years post-baseline. These data will cover caregiving activities, physical effort of caregiving, depressive symptoms, stress, coping, stress-related behaviors, and weight and ADL/IADL functioning. These data will be linked to the 6th SOF clinic data on physical and functional health, cognitive status, weight, physical activities, sleep disturbance, and neuromuscular functioning, and SOF follow-up data on falls, fractures, heath care use, and mortality. Longitudinal analyses will test hypotheses that caregivers have a higher rate of health decline than non-caregivers (e.g., change in ADLs/IADLs, falls, fractures, weight change) and hypotheses on the associations among caregiving activities, stress-related behaviors and health decline among white versus African-American caregivers. This study will benefit from its prospective design, its link to a large multi-site population-based cohort of elderly women, and a wealth of standardized self-report, clinical, and functional measures on physical health. The number of elderly caregivers in the United States is growing; this group is at highest risk of adverse health effects of caregiving. These study results will provide epidemiologic evidence of adverse physical health outcomes of caregiving, and implications for detection and prevention of health decline in caregivers.